Who this checklist is for (and where it stops)

I'm the procurement manager at a 150-person rice trading and processing company. For the past seven years I've controlled a capital equipment budget of roughly $900,000 a year, and in that time I've signed off on just over 30 rice milling machine purchases — single commercial rice milling machines for our satellite buying points, plus two full industrial rice milling lines, the larger one commissioned in 2022.

What follows is the seven-step checklist I run before any PO goes out. It's built for buyers adding or replacing capacity in the 15–120 t/day range. If you're tendering a 300 t/day plant, some of it still applies, but your process engineering is a bigger job than mine.

Seven steps, in the order I actually do them. Straight through, it takes me about three weeks. I've compressed it into five days once. The step I skipped was step 7, and it cost us nine days of idle equipment — worse than nine days of patient waiting.

The 7-step sourcing checklist

1. Convert your real throughput into the number you put in the RFQ

Brochure capacity is paddy input, not finished rice out, and it's measured under conditions you probably don't have. A machine quoted at 20 t/day means 20 tonnes of paddy per 24 hours at around 14% moisture, with clean grain and zero stoppages.

Our wet-season paddy comes in at 18–20%. When that's the raw material, I plan on roughly 0.8× the nameplate and treat anything above that as a good day. If a supplier won't tell you the moisture and cleanliness assumptions behind their number, that's your first data point about the supplier.

Two things to settle before you size anything: your shift pattern, and your head rice yield target. A 5 t/h machine on one shift isn't the same asset as the same machine running 20 hours. On the mid-size lines we run, I budget 55–62% head rice from paddy depending on variety and moisture. Your miller will have sharper numbers than I do — ask them, not me.

2. Lock the spec sheet before you ask for a price

Price without a spec is a number you can't compare to anything. Ours covers, at minimum:

  • Capacity in t/h of paddy, stated at a defined moisture level
  • Paddy moisture range the machine will handle without choking
  • Husker type and roller dimensions (rubber roller wear is a real line item)
  • Whether a destoner, separator, and polisher are inside the quoted scope
  • Power supply: 380V/50Hz three-phase or 440V/60Hz, and total connected load in kW
  • Which wear parts ship with the machine, and which you buy separately

Colour sorters are usually quoted as a separate line, and that's a big number — commonly $20,000–$60,000 depending on channels and throughput. Decide up front whether sorting is in this budget or the next one, because it changes the layout and the power draw.

3. Compare landed cost, not machine price

This is where most of the money lives. In 2023 I had two quotes side by side for what was nominally the same 40 t/day commercial rice milling machine. Supplier A: $31,900. Supplier B: $28,400. On paper, B wins by $3,500.

Then I rebuilt both into the same total-cost line and the gap flipped. B's quote was FOB, so we'd have paid $2,900 in freight. The destoner wasn't included — $4,200 to add. No commissioning engineer, which meant $3,100 plus flights and hotels for ours. No spare roller set. Real total: about $38,600. A's price included freight to port, the destoner, five days of commissioning, and one spare roller set: $33,400 landed and running. Roughly a 15% swing that existed only in the fine print.

My column headings are: machine, freight and insurance, duty and clearance, installation and commissioning, electrical works, training, first spare parts order, and expected wear-part spend for year one. Every quote gets rebuilt into that grid before it goes to my director.

4. Verify what the vendor actually makes — and listen when they say 'we don't make that'

Plenty of rice milling machine suppliers will tell you they build everything: the mill, the dryer, the colour sorter, the bagging line, the steel structure. Some of them genuinely do. Many of them assemble and buy in more than they admit.

Ask directly which components are bought in, and from whom. Motors, bearings, blowers, and the pneumatic roller adjustment are the four I always chase, because those are the parts that decide whether you're down for four hours or four weeks.

The vendor who told me their dryer line was bought in, and pointed me to a specialist for it, got the milling order anyway. A supplier who says 'this isn't our strength' has just told you their other answers are calibrated. That's worth more than a brochure.

I'd rather buy a milling machine from someone who knows exactly where their competence ends than from someone who claims a one-stop shop and then subcontracts the hard half.

5. Call references that match your scale, and ask uncomfortable questions

Two or three references, in the same capacity band, running the same raw material. Reference lists full of installations twice your size aren't much help — a 200 t/day plant has maintenance staff who absorb problems yours can't.

The questions that actually produce information, in my experience:

  • How many hours a month was the line down in the last six months, and why?
  • When you needed a wear part, how long did it take to arrive?
  • Who came out for the last service call — the manufacturer's own technician or a local contractor?
  • How many days did commissioning take against the plan?

Anyone who answers those four with specifics is a reference worth having. Anyone who answers with praise and no numbers is a name on a list.

6. Put the boring contract terms in writing

Four clauses do most of the work: warranty scope, spare parts availability, training, and the acceptance test.

Split the warranty, because lumping it together hides where you're exposed. Structural and main-frame coverage of 12–24 months is normal; wear parts run 3–6 months, and rubber rollers are sometimes warranted by tonnage milled rather than by calendar. Get the split in the contract, not in an email.

Ask for written spare parts availability of 7–10 years for the model you're buying. Ask how many days of on-site training are included and whether that's classroom or at the control panel. And tie the final payment to a performance test at rated capacity, on your paddy, at your moisture. If a supplier won't accept a capacity-based acceptance test, that tells you what they think of their own nameplate.

7. Get the site ready before the machine ships — the step everyone skips

This one isn't about the supplier at all, which is exactly why it gets dropped. While the machine is on the water, you need to have settled:

  • Foundation and civil work, including anchor points and drainage
  • Transformer and panel sizing against connected load, plus cable runs
  • Dust extraction and ventilation — rice dust is a real problem, not a cosmetic one
  • Discharge and bagging routing, so finished rice has somewhere to go
  • Access for the container and the crane on delivery day

In 2023 we had 11 days between the machine clearing port and harvest starting. Everything was ready except the incoming transformer capacity. We lost nine working days with the line installed and unpowered. The machine was fine. We weren't.

Six mistakes that cost us real money

  • Comparing quotes with different scopes. The single most common error, and it usually flatters the worst supplier, because the worst supplier leaves things out.
  • Budgeting the machine and forgetting the landing costs. Freight, duty, installation, and commissioning add 12–20% on top of the FOB price in my records. Budget for it or get surprised by it.
  • Ignoring wear-part consumption. On our 60 t/day line we go through a set of husker rollers roughly every 350–450 running hours, if I'm remembering the log correctly. That's a recurring cost, not a one-off.
  • Not asking the spare parts lead time. A $900 part with a 10-week lead time is really a 10-week outage with a $900 invoice attached.
  • Skipping the performance test. Run it at rated capacity on your own paddy. Machines that pass on dry, clean grain sometimes fail on what you actually buy.
  • Assuming the electrics match. A 50Hz machine on a 60Hz supply spins about 20% faster. That's not a small installation detail; that's a warranty conversation.

Where my numbers stop being useful

I'm not a milling engineer. I can't tell you the right roller pressure, the ideal screen perforation, or how to tune aspiration on your specific variety — that's a process engineer's call, and I'd be guessing. What I can tell you is how to make sure the commercial and contractual side doesn't quietly cost you 15%.

My sample is also narrower than it sounds. Thirty-odd purchases, mostly commercial rice milling machines in the 15–120 t/day band, all sourced internationally but installed in similar operating conditions. If you're running a very small village-scale mill or a fully automated 300 t/day industrial plant, the weights shift and some of this won't transfer cleanly.

Price ranges in this article are drawn from quotes I've seen between 2023 and 2025: roughly $9,000–$35,000 for a single 15–25 t/day commercial rice milling machine, $75,000–$180,000 for a complete 50–80 t/day line, and $350,000 upward for industrial-scale installations with sorting and automation. Treat those as sanity checks for your own budgeting, not as quotes. Pricing varies by configuration, port, and month — verify current rates before you commit.

If you only take one thing from this: the cheapest quote I've ever received was the most expensive machine we ever bought. Not because it broke, but because of everything that wasn't in the price.